This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Using NLS in the regression checker
On Sat, Oct 21, 2000 at 08:49:55PM +0200, Philipp Thomas wrote:
> * Zack Weinberg (zackw@stanford.edu) [20001021 19:49]:
> > This doesn't help with *added* files, any idea how to detect them?
>
> At least no easy way. You'ld need to scan the whole gcc tree for .c
> or .h files that contain one of the keywords (direct gettext calls,
> the _() or N_() macros or functions that have a parameter named
> msgid), remove those files whose messages should never be translated
> (like the gen* tools or sources for libgcc), make the list relative
> to the gcc directory (so that the list is the same format as
> POTFILES.in) and then check if an entry in this list is also present
> in POTFILES.in.
The convention at the moment seems to be that every file is in the
list, whether or not it has translatable messages in it. This seems
sensible to me - you don't want to worry about updating POTFILES.in
every time you edit a file. As such, it would be easy enough to write
a shell script that takes an exclude list and the output of
'find . -name "*.[ch]"', and spits out a POTFILES.in.
Maybe I'll do just that later today.
zw